This comprehensive manual is designed to give archaeologists the necessary background knowledge in environmental science...
123 downloads
1260 Views
23MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
This comprehensive manual is designed to give archaeologists the necessary background knowledge in environmental science required to excavate and analyse archaeological sites by rivers and on floodplains. Part I covers the techniques for studying alluvial environments, while Part II reviews the literature on the archaeology of alluvial environments. An important theme running through the book is the interaction between climatic and cultural forces and the transformation of riverine environments. Bringing together information on the evolution and exploitation of floodplain and river landscapes, it draws on examples from Britain, Europe, North America and Australasia. Alluvial geoarchaeology will also interest physical geographers, geologists and environmental scientists.
ALLUVIAL GEOARCHAEOLOGY
CAMBRIDGE MANUALS IN ARCHAEOLOGY Series editors Don Brothwell, University of York Graeme Barker, University of Leicester Dena Dincauze, University of Massachusetts, Amherst Priscilla Renouf, Memorial University of Newfoundland Already published J. D. Richards and N. S. Ryan, DATA PROCESSING IN ARCHAEOLOGY Simon Hillson, TEETH Alwyne Wheeler and Andrew K. G. Jones, FISHES Lesley Adkins and Roy Adkins, ARCHAEOLOGICAL ILLUSTRATION Marie-Agnes Courty, Paul Goldberg and Richard MacPhail, SOILS AND MICROMORPHOLOGY IN ARCHAEOLOGY
Clive Orton, Paul Tyers and Alan Vince, POTTERY IN ARCHAEOLOGY R. Lee Lyman, VERTEBRATE TAPHONOMY Peter G. Dorrell, PHOTOGRAPHY IN ARCHAEOLOGY AND CONSERVATION (2ND EDITION)
Cambridge Manuals in Archaeology are reference handbooks designed for an international audience of professional archaeologists and archaeological scientists in universities, museums, research laboratories, field units, and the public service. Each book includes a survey of current archaeological practice alongside essential reference material on contemporary techniques and methodology.
ALLUVIAL GEOARCHAEOLOGY Floodplain archaeology and environmental change
A. G. Brown University of Exeter
CAMBRIDGE UNIVERSITY PRESS
CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo Cambridge University Press The Edinburgh Building, Cambridge CB2 2RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521560979 © Cambridge University Press 1997 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1997 Reprinted 2001 A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Brown, A. G. Alluvial geoarchaeology: floodplain archaeology and environmental change / A. G. Brown. p. cm.-(Cambridge manuals in archaeology.) Includes bibliographical references (p. ) and indexes. ISBN 0 521 56097 7 - ISBN 0 521 56820 X (pbk) 1. Excavations (Archaeology)—Handbooks, manuals, etc. 2. Watersaturated sites (Archaeology)—Handbooks, manuals, etc. 3. Soil— Analysis-Handbooks, manuals, etc. 4. Floodplain ecologyMethodology-Handbooks, manuals, etc. 5. Archaeological geologyMethodology-Handbooks, manuals, etc. I. Title. II. Series. CC77.5.B76 1997 930.1'028-dc20 96-2413 CIP ISBN-13 978-0-521-56097-9 hardback ISBN-10 0-521-56097-7 hardback ISBN-13 978-0-521-56820-3 paperback ISBN-10 0-521-56820-X paperback Transferred to digital printing 2006
To Sara and Gabriel
CONTENTS
List of illustrations List of tables Preface Acknowledgements Introduction and the example of the Nile
page xii xx xxi xxiii 1
PART I P R I N C I P L E S 1
Flood plain evolution 1.1 Floodplain evolution: an introduction 1.2 River channel change and floodplains 1.3 Aggradation and degradation 1.4 Terraces 1.5 Floodplain evolution and archaeological sites 1.6 Sites and rates of sedimentation 1.7 Alluvial geoprospecting and site conservation
17 17 25 33 34 37 39 41
2
Alluvial environments over time 2.1 Dating and alluvial chronologies 2.2 Chronometric dating methods 2.3 Incremental and calibrated relative techniques 2.4 Archaeological dating: structures and pots in alluvium 2.5 Stability and change
45 45 48 55
3
Interpretingfloodplainsediments and soils 3.1 River morphology and sedimentation 3.2 Floodplain sediments 3.3 Palaeohydrology 3.4 The fluvial transport of archaeological materials 3.5 Floodplain soils 3.6 Archaeology and floodplain palaeosols
IX
58 60 63 63 70 81 91 96 100
Contents 4
Floodplain ecology, archaeobotany and archaeozoology 4.1 Floodplain productivity 4.2 Hydrological controls on floodplain vegetation 4.3 Vegetation and sedimentary environments 4.4 Colonisation and succession 4.5 The floodplain and river continuum 4.6 Ecological studies of floodplains in different environments 4.7 Floodplain palaeoecology 4.8 Dendroclimatology
104 104 104 108 115 117 118 128 145
PART II APPLICATION 5
Artifacts fromfloodplainsand rivers 5.1 Palaeolithic terrace sites and floodplain use 5.2 Geoarchaeological studies of terrace gravels in southern England 5.3 North American alluvial geoarchaeology 5.4 Australasian alluvial geoarchaeology 5.5 Geoarchaeological studies and archaeological inference 5.6 Interpreting artifacts from rivers
150 167 184 189 190
6
The rise and fall of forestedfloodplainsin North-West Europe 6.1 Lateglacial fluvial change and the Upper Palaeolithic 6.2 Early Holocene fluvial environments and the Mesolithic 6.3 Mid-Holocene floodplain forests and human impact 6.4 Floodplain deforestation
192 193 199 210 215
7
Buried sites 7.1 Late Bronze Age and Iron Age alluviation in the British Isles 7.2 Roman and medieval alluviation and channel change in the British Isles 7.3 Historical channel change and alluviation in mainland Europe 7.4 The Mediterranean record 7.5 The North American experience 7.6 The role of climatic change
219
8
Managed floodplains 8.1 Flood farming and irrigation 8.2 Temperate flood-meadows 8.3 River management: from early obstructions to wellbehaved rivers
149 149
219 225 235 237 248 251 254 254 256 257
Contents
xi
8.4 Land drainage 8.5 The reconstruction of historical flood magnitude and frequency
268
The cultural archaeology of floodplains 9.1 Locational data and alluvial environments 9.2 Positive factors 9.3 Negative factors 9.4 Locational synthesis
279 279 281 291 300
10 People,floodplainsand environmental change 10.1 Palaeohydrology, climate and resources 10.2 Floodplain settlement and environmental change: past and future
304 304
9
APPENDICES Notation 1 River flow and sediment transport 1.1 Flow 1.2 Sediment transport 1.3 Grain size analysis and palaeohydraulics
270
313 318 320 320 322 327
2
Flood frequency analysis 2.1 Standard flood frequency analysis 2.2 Historical flood frequency analysis
331 331 332
3
Documentary evidence and wetland perceptions 3.1 Place-and river-name evidence 3.2 Literary evidence and the human perception of wetlands
333 333 335
References Subject Index Index of rivers and sites
338 372 375
ILLUSTRATIONS
Plates 1.1
1.2 2.1 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12
An aerial photograph of the junction of the rivers Trent and Soar, England, under flood conditions, highlighting floodplain topography associated with meander migration and avulsion. A small floodplain from near Antequera in southern Spain showing the stripping out of much of the floodplain caused by a single rain event in November 1991. The excavated block from Cossington, Soar valley, England, from which radiocarbon samples were taken. A braided, wandering gravel-bedded river in north-eastern England. Secondary flow circulation caused by bank irregularities and picked out by foam, on the river Liffey, Ireland. Dead water with a secondary flow vortex picked out by foam at the junction of a backwater and a main channel, from the river Ae, Dumfriesshire, Scotland. A section through the drained bed of the river Perry, Shropshire, England. Point-bars developing from alternating bars on a river in eastern Spain. Mudballs on the floodplain of the river Turon in southern Spain after a flood in November 1989. Levee sediments formed by a single flood of the river Soar, England. Laminated overbank sediments from the Lower Severn, England. A trash line at the crest of an artificial levee of the river Severn, England. Erosion of overbank deposits from a flood of the river Turon in southern Spain, November 1989. Erosion of the floodplain soil after a flood on the river Ae, Dumfriesshire, Scotland. Rills, erosion and colluvial deposits at the edge of the
xii
28 32 52 66 68 69 71 74 75 76 77 77 78 78
List of illustrations
3.13 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 5.1 5.2 6.1 6.2 7.1 7.2 7.3 7.4 7.5 7.6 8.1 8.2
floodplain and slope base, caused by a single storm, Leicestershire, England. A section of the upper palaeosol from the river Perry, Shropshire, England, revealing a c. 6800 year old burrow. The Ae in Dumfriesshire, Scotland, has reaches that are meandering, stable, braided and anastomosing. Lobes of bedload gravel splaying out from the channel and deposited on an older bar surface on the river Ae, Dumfriesshire, Scotland. A small debris dam from the New Forest, England. Mid-channel bar ridges formed behind fallen and transported trees on the river Ae, Dumfriesshire, Scotland. A beaver dam and lodge from a small Polish river. The Gearagh, Co. Cork, Ireland: islands inundated in a flood. The Gearagh, Co. Cork, Ireland: discharge across an islet. Channels within the Litovelske Pomoravi wooded floodplain, Czech Republic. Palaeochannels on the Riverine Plain, south-eastern Australia. Palaeochannels on the Riverine Plain, south-eastern Australia. Lateglacial peat lens from an eroded channel fill from the Raunds reach of the Nene floodplain, Northamptonshire, England. An in situ root cluster of alder (Alnus glutinosa) from the Raunds reach of the Nene floodplain, Northamptonshire, England. The Buff-Red Silty-Clay of the Severn valley, exposed at Powick near Worcester, England. An aligned oak (Quercus) trunk from Colwick in England. The excavated remains of an eleventh-century bridge across the river Trent, England. The alluvial stratigraphy of the river Werra in north Germany. A typical 'Younger Fill' type alluvial unit from the Marta valley, near Tarquinia in southern Etruria, Italy. A 4 m accumulation of banded sand and silt adjacent to and partly burying a Roman structure. The remains of a medieval (Norman) fish weir excavated from the Trent valley at Colwick, England. Holme Fen post, Fenland England, showing the shrinkage of the peat since 1848.
xiii 83 103 109 109 113 114 114 125 126 127 186 186 195 218 225 232 234 237 245 245 260 271
xiv
List of illustrations 8.3 8.4 9.1 9.2 9.3 9.4
The Ripetta pillar marking flood heights of the river Tiber in Rome. Flood heights on the Water Gate of Worcester Cathedral on the banks of the river Severn, England. A flood door on a seventeenth-century cottage on the floodplain of the river Axe in Devon, England. A flood walkway, near the village of Sutton Bonnington, Leicestershire, England. The Ferry Boat Inn at Stoke Bardolph, Nottinghamshire, England. The floodplain at Gundagai, New South Wales.
272 273 296 297 298 300
Figures Intro. 1 Hapi the Nile God. Intro.2 Map of the Nile catchment and Egyptian Nile valley showing major geomorphological features and localities mentioned in the text. Intro.3 Simplified Nile valley cross-section near Tahta. Intro.4 Height of Nile floods in metres above an arbitrary datum and calibrated 14C for increased lake volumes in the Sahara and East Africa since 5000 BC. Intro. 5 The climate-culture chain of the Nile valley. 1.1 Block diagram of landforms associated with a meandering river and its floodplain. 1.2 Idealised and simplified floodplain sedimentary systems. 1.3 The stable-bed aggrading-banks (SBAB) model of floodplain and channel evolution. 1.4 River channel planform and planform changes. 1.5 A chute cutoff and its generalised grain size distributions from point-bar, concave-bank bench and the fill. 1.6 Stratigraphy produced by the avulsion model of Bridge and Leeder(1979). 1.7 A diagrammatic representation of the disequilibrium stripping model of Nanson (1986). 1.8 Terrace morphology in relation to long-profile changes and combinations of terraces with alluvial fills. 1.9 The effects of alluvial sedimentation rate on artifact densities and spatial patterning. 1.10 The effects of bioturbation illustrated using grain size curves from non-cultural deposits and cultural deposits around the Carlston Annis mound, Kentucky.
6 8 9 11 13 18 22 24 27 29 30 31 35 36 40
List of illustrations 1.11 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15
The generalised modern stratigraphy, artifact distribution and dating of a cross-section of the Duck river floodplain, Tennessee. Time-spans covered by different dating methods used in alluvial environments. Time-lines in different geomorphological systems. Block diagram of radiocarbon dated infill of a palaeochannel from the floodplain of the river Soar at Cossington, England, excavated by M. Keough. Diagram illustrating the different geomorphological significance of four radiocarbon dates from three different stratigraphies. Deposition of dated tree trunks (rannen) over time in the Main and Danube rivers. Acid racemisation values and a tentitive correlation of the Avon terraces, England, and marine terraces. Factors in the inclusion of pottery into alluvial sediments Models of landscape change over time. Definition sketch of a step pool channel, (a) a pool-riffle sequence (b), and the relationship between velocity/shear stress and discharge in a riffle-pool section (c). Morphological classification of channel bars. Channel pattern types. Discrimination between straight, meandering and braided channels. Flow circulation around a meander bend. The Keller five-stage model of meander development. Sediment fabric, structure and architecture. Typical point-bar sedimentary architecture. Accumulation rates from radiocarbon dated units from the floodplain of the river Severn, England. CM diagram. Flood deposits illustrating the relationship between event and stratigraphy at two different heights and a flood frequency curve from the Salt river in Arizona. The flood stratigraphy of the Columbia river site, Washington State. Overloose, normally loose, armoured and cemented channel beds. The planform and morphology of meanders of the Posna river, Poland, used for palaeohydrological reconstruction. Forces acting on a submerged bone.
xv
42 46 47 51 53 56 59 60 62 64 64 65 66 68 70 72 73 80 82 85 85 88 90 93
xvi
List of illustrations 3.16 3.17 3.18 3.19 4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18 4.19 5.1 5.2 5.3
Predicted transportability against observed transport distance for selected bones. The two piles of mammoth bones in the arroyo at Colby, Kansas. The mean residence time of soil in the pedogenic zone of an alluvial sequence and its probable effect on vertical artifact distribution. An exposure of the palaeosols in the floodplain sediments of the river Perry, Shropshire, England. Theoretical floodplain and non-floodplain annual productivity and the differential productivity with climate. A simplified representation of the hydrology of a temperate floodplain. Floodplain watertable rise and the capillary fringe (a) prior to rainfall and (b) after rainfall. Channel-edge ecotones and the vegetation gradient. The age classes of cottonwood trees on the Little Missouri floodplain, Wyoming. The river continuum model and the relative importance of functional feeding groups and land-water ecotones. A vegetation transect across the Tanana river valley, Alaska. A vegetation transect across the Lower Amazon floodplain. Vegetation classes from the Rio Grande floodplain. A vegetation transect across the Sacremento valley in California. Vegetation associations across the river Dyje floodplain in Czechoslovakia. Synchronic and diachronic analysis. The chain of inference in palaeoecology. A simplified outline of the Troels-Smith method of recording organic sediments. The internal structure of wood showing the sections used for identification and a tree-ring curve. The Jacobsen and Bradshaw model of pollen recruitment. A modified version of the Tauber model for a small oxbow lake with a small stream input. The decline in lime pollen across a floodplain. Dendrohydrological reconstruction of streamflow from northern Arizona back to 1580. A schema of terrace development for the Thames, England. A map of the Thames valley showing the location of major sites mentioned in the text. A diagrammatic summary of the artifact-bearing units in the Upper Thames.
94 95 100 102 105 106 107 110 116 118 119 120 122 123 124 128 130 132 133 137 138 140 146 151 155 158
5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15 5.16 5.17 5.18 5.19 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8
List of illustrations
xvii
The diverted Anglian course of the Middle Thames. Detail of part of the east face of Warrens Lake Pit, Yiewsley, Thames valley. Generalised stratigraphy from Swanscombe, Thames valley. Map of North America with regions and sites mentioned in the text. Kersey site, Dent County, and the Kersey terraces, Colorado. Laddie Creek stratigraphy, Wyoming. The Dyer site, W. Oklahoma. The stratigraphy at Plainview, Texas. A generalised cross-section of the alluvial stratigraphy of Whitewater Draw, Arizona. Generalised stratigraphy from the northern section of the San Xavier reach of the Santa Cruz river, Arizona. Landscape reconstruction and settlement patterns from the San Xavier reach of the Santa Cruz river, Arizona. Palaeovegetation maps for eastern North America, 14,000, 10,000 and 200 bp. The five meander belt stages of the Holocene development of the Central Mississippi. The distribution of Mississippian sites on the north bank of the Ohio in Black Bottom, Ohio. The Murray-Darling catchment and palaeochannels of the Riverine Plain around Hay, New South Wales. The evolution of the environment of the Kakadu region, Northern Territories, Australia. Location of sites in the East Midlands, England, mentioned in the text. Stratigraphy of the site at Ditchford in the Nene valley, England. A representation of the general stratigraphy at Sproughton on the river Gipping, England, and location of the barbed points. Time correlations of the assemblage zones from the valley cores and the upland pingo cores in the southern Netherlands. An interpretive reconstruction of the depositional sequence at Star Carr, Yorkshire, England. Map of the Shippea Hill sites with an outline of the stratigraphy at Plantation Farm, Cambridgeshire, England. A reconstruction of the fen edge environment from the channel at Peacock's Farm, Cambridgeshire, England. A sketch map and part of the stratigraphic sequence at Thatcham, Berkshire, England.
160 161 163 168 171 172 173 174 176 177 179 180 181 181 185 188 194 196 198 201 202 204 205 207
xviii
List of illustrations 6.9
6.10 6.11 6.12 6.13 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 7.10 7.11 7.12 7.13 7.14 7.15 8.1 8.2 8.3 8.4 8.5 8.6
Reconstruction of the environment around the Mesolithic site of Noyen-sur-Seine, Paris basin, France. Map of the Pleszow settlement in the Vistula valley, Poland. A visualisation of the settlement phases at Pleszow, Poland. Plans of alder woods derived from in situ tree roots. The alder decline in the Severn valley, England. Runnymede: the Late Bronze Age river edge with structures. Floodplain accretion rates and estimated erosion rates for the Ripple Brook catchment, Worcestershire, England. Historical river channel changes in the Lower Severn basin, England. Historical channel changes from the Middle Severn at Welshpool, Wales. Channel changes on the river Dane at Swettenham, Cheshire, England. Medieval channel changes on the river Trent at Colwick, near Nottingham, England. A generalised cross-section from the Upper Main, in Germany. Cross-sections of the Leine, lime and Weser valleys, northern Germany, and magnetic susceptibility profiles. Terrace sequence of the Voidomatis basin, north-west Greece. A map of central Italy showing rivers and sites mentioned in the text. The alluvial cross-section at Narce, Etruria, Italy, and its interpretation. Longitudinal section of the Torre di Lama, Foggia, southern Italy. Generalised average floodplain logs from the lower reaches of five valleys in southern Etruria, Italy. The history of alluviation, vegetation and culture in the southern Argolid, Greece. Sediment budget studies in Coon Creek, Wisconsin. The alluvial fan, discontinuous channel system and optimum location for Ak-Chin farming. Part of the floated meadow system at Lower Woodford in the Avon valley near Salisbury, England. An excavated Saxon fish weir from Colwick in the Trent valley near Nottingham, England. A map of medieval fish weirs on the Middle Severn, England. Five generalised types of mill and leat plans. Litovelske Pomoravi, Czech Republic.
209 212 212 216 217 220 224 228 229 230 233 236 238 240 241 242 243 244 246 250 255 258 259 261 262 264
List of illustrations 8.7 8.8 8.9 8.10 8.11 8.12 8.13 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 10.1 10.2 10.3 10.4 10.5 Al A2
A generalised 1944 and 1973 cross-section across the Middle Mississippi floodplain with the maximum containable flood in 1973. A map of the eighteenth-century shoals in the river Severn between Gloucester and Stourport, England. A combined map of the channelised rivers of England and areas of predominantly wetland prior to this century. Flood maps from Florence, Italy. The frequency of levels taken at Coalport, Shropshire, England, 1789-1800. The reconstructed single-peak, in-bank hydrographs from Coalport, 1789-1800, and Ironbridge 1951-62. Non-dimensional curtailed flow duration curves for the Coalport data (1789-1800) and the Ironbridge data (1951-62). The preservation potential of a hypothetical floodplain. The distribution of log-boat finds in England and Wales. Spatial efficiency of the Mississippi Moundsville sites, West Virginia. Generalised flood frequency curves. A generalised graph of floodplain risk in relation to flood frequency and magnitude. Map of the old and new north town of Gundagai, New South Wales. Settlement patterns in the Lower Mississippi valley. The chain of factors between hazard/resource and behaviour. The reasoning chain in palaeohydrology. A diagram illustrating the nested frequency-magnitude concept. A simple rational model of floodplain dwellers' response to flood hazard. A schematic representation of the cascading pyramid of uncertainties associated with the global warming problem. A hierarchy of models of climate-society relationships. A diagram of channel geometry and hydraulic parameters. A modified Hjulstrom curve.
xix
266 267 268 274 276 276 277 280 285 290 292 293 299 301 302 305 308 309 314 316 321 324
TABLES
Intro. 1 The components of geoarchaeology. 1.1 The sedimentary features and mode of formation of the major micro-landforms found on floodplains. 3.1 Computed discharges for the Washington river for AD 120 to 1948. 3.2 Commonly used formulae for the computation of palaeodischarges. 3.3 The classification of alluvial soils in the UK system. 3.4 Some common minerals formed in floodplain sediments. 3.5 Studies of soil chronosequences on alluvial landsurfaces. 4.1 The general relationship between vegetation associations and floodplain topography for bottomland hardwood forests in south-eastern USA. 4.2 The condition of pollen grains. 4.3 The typical size, specific gravity and terminal velocity of selected pollen grains. 4.4 The ecological groups of freshwater snails. 5.1 A simplified Pleistocene chronology for southern England with climate and, where possible, estimates of sea level. 5.2 The Thames terrace and channel members. 8.1 Recorded floods from Florence, 1100-1899. 8.2 Unqualified Severn floods. 9.1 A list of some of the plants known to be used as food or poison from British floodplains. 10.1 Pay-off matrices for floods: expected utility and known probability. Al Grain size scale and ranges of different techniques.
xx
2 20 86 87 97 99 101 111 136 139 143 154 156 275 275 284 310 329
PREFACE
The origins of this book may help to explain its existence, structure and content. Having undertaken doctoral research jointly supervised by a geomorphologist and a palaeoecologist which had strong archaeological implications, I was by the mid-1980s firmly situated in a multidisciplinary and interdisciplinary approach to fluvial environments. Whilst at Leicester University my contacts grew with researchers in other departments interested in alluvial environments, most notably Zoology, Botany and Archaeology. Increasing contact with archaeologists both at Leicester University and in Leicestershire and Northamptonshire Archaeological Units stimulated my interest in geoarchaeology and the cross-fertilisation of geomorphology and archaeology. This was given a research foundation when I realised that archaeologists were regularly digging large holes in floodplains and valued stratigraphic assistance. I worked first on the Raunds Area Project, Northamptonshire, and since then have worked on many alluvial sites in the Midlands of England and elsewhere. In 1989 I took a joint appointment between Geography and Archaeology at Leicester and began to teach environmental archaeology, which included geomorphology, to archaeologists. Contact with archaeologists on a daily basis has I hope benefited this book, not least through the undermining of the simplistic and naive tendencies concerning culture and society that are all too typically held by natural scientists. This book stems from this history and a desire to bring together work in a variety of disciplines pertinent to the study of alluvial archaeological sites and floodplain geomorphology and palaeoecology. The book was technically conceived on a train journey back to Leicester from the Institute of Archaeology in 1985 but as this is a short journey it took some years before writing began in earnest. Its writing was given further impetus by two factors: first the very limited and generally simplistic treatment of floodplain forms and processes in most textbooks, even advanced books, and secondly a need for students and researchers to have a source of multidisciplinary information on recent and contemporary alluvial environments. There are probably no processes on floodplains which are not in reality influenced by both physics and biochemistry (in the widest sense) and in most floodplains are simultaneously artifacts of human activity. It is extremely difficult for any scientist (an archaeologist in particular given the typical archaeological training) to find an entrance into cognate disciplines which may have great importance for the understanding of
xxi
xxii
Preface
processes on floodplains. The compartmentalisation of knowledge into rather arbitrary 'disciplines' which has been given renewed impetus by recent economic and academic pressures is fundamentally counterproductive in subjects such as this. Whilst the book is primarily designed for archaeologists, geomorphologists and palaeoecologists it is hoped that it will also be of interest to others, in particular engineers and ecologists, involved in current and future attempts to improve our environment by the restoration of 'natural' floodplains. Most of all I hope that this book exemplifies a firm belief of the author's, which is that, as in science and the arts, new knowledge, at least initially, springs from the association of previously unassociated phenomena.
ACKNOWLEDGEMENTS
To acknowledge all who have influenced this book would be impossible as its conception dates back several years and it draws on all who have influenced me from an indefinable date. There are, however, several groups of individuals and institutions that have given tangible help and are acknowledged here. Many of the ideas and case studies in the book come from discussions with my colleagues at Leicester University. In particular I must thank Rob Young, Graeme Barker and John Rice for reading and commenting on earlier drafts. My work, and this book, has greatly benefited from association with members of the former Leicestershire Field Archaeology Unit, particularly Patrick Clay, and other field archaeologists including Chris Salisbury and Ian Meadows. Many archaeologists allowed me access to unpublished or in press material for which I thank them. I have drawn on geomorphologists who have run field visits, including Martin Thorpe, Mark Macklin, Jim Rose and Jeff Vandenberghe, for which I thank them, and on the intellectual background of the British Geomorphological Research Group and the Quaternary Research Association, and also my original mentors, K. J. Gregory, K. E. Barber and C. Vita-Finzi. I must also thank members of the Institute of Archaeology, London University, for inadvertently providing the initial impetus for the book. The production of this book could not have been possible without the superb drawing skills of Kate Moore, Ruth Pocklington and Sylvia Holyoak at Leicester and Terry Bacon at Exeter. I must also thank the Central Photographic Unit at Leicester and Andrew Teed at Exeter for excellent photographic services. Comments made on an earlier draft by at least three anonymous referees were, I believe, extremely useful in improving the whole book. I thank all those who have allowed me to use their work, and take full responsibility for errors of interpretation and fact. The prolonged gestation of this book has been made much easier by the efficient and helpful editorial support of Jessica Kuper. Proof reading was facilitated by the kindness and generosity of the archaeologists of the Rudi-Maetonium Project, Moldova. Lastly but certainly not least I must thank Sara Mills for her attempt to force stylistic improvements and for tolerance of the whole enterprise and G. Mills-Brown for not eating all the disks.
xxin
INTRODUCTION AND THE EXAMPLE OF THE NILE
What makes riverine or 'alluvial' environments different, both from other environments and from each other, and how does this affect the archaeological record? How can we study environmental change in alluvial environments and what impact has it had on human populations? This book aims to answer these questions. It also aims to provide an introduction to the physical and biological aspects of alluvial environments which are central to an understanding of archaeology on, under and near floodplains. Questions of preservation, transportation, burial, environment and subsistence are all intimately related to the characteristics of the landscape and are also essential components in any archaeological interpretation. Another aim of the book is to introduce archaeological aspects of alluvial history to environmental scientists and geographers because the vast majority of, if not all, contemporary floodplains have to a greater or lesser degree been altered by human activity during the last 10,000 years. Indeed some have been so altered as to make them in part artifacts and as such indicators of the impact of humans on the environment. This book is therefore about both the impact of humans on their environment, and the impact of the environment on humans. In order to illustrate this and lead the reader through the complete cycle of the inference of cultural implications from the environmental data a classic example is used: the Nile. It is all too easy for specialists, be they archaeologists or geomorphologists, to work as part of a project team on a particular area or problem and never get to see the 'big picture'. This introductory chapter and the rest of the book, whilst promoting specialist analysis of environmental data, also tries to stress the importance of a wider awareness of the potential uses and abuses of such data. It is abuse of the data which has in the past led to the twin evils of environmental relativism (i.e. the view that the environment does not have any particular role to play) and environmental determinism (i.e. the view that environment determines human behaviour). In the last few years there has been an explosion of interest in the archaeology of wetland and alluvial environments at the research level (Coles, 1992a; Needham and Macklin, 1992). This raises the question of whether there is any such thing as the archaeology of a physical environment such as coast, floodplains or mountains. The archaeology of a particular physical environment may be distinctive for two reasons. First, because much cultural history and all prehistoric culture can only be viewed through the physical and biological 1
2
A lluvial geoarchaeology
Table Intro. 1 The components of geoarchaeology, adapted from Hassan (1979) andGoudie (1987). Components 1 site location 2 geomorphological analysis of site environs 3 regional stratigraphic studies 4 sedimentary analysis of deposit 5 palaeoenvironmental analysis 6 modelling relationships between human activities and landscape 7 studies of natural hazards 8 dating
Typical methods/data topographic maps, thematic maps, remote sensing, geographical information systems field mapping, stratigraphy, dating collation of geomorphological studies, remote sensing facies identification, studies of process and provenance, e.g. mineralogy, texture, mineral magnetics facies interpretation,palaeoecology, e.g. snails, pollen, wood, phytoliths, diatoms, insects correlation of environmental and cultural change, resource analysis, catchment area analysis, carrying capacity and Malthusian models most of the above radio-isotopes, luminescence dating, acid-racemisation, palaeomagnetic dating
remains of material culture. Therefore, irrespective of more elaborate theorising, the first step is the reconstruction of the material base of societies. The conditions of burial, which control the degree and bias of preservation, are therefore of fundamental importance for all archaeologists (Clarke, 1973). Like other natural environments, alluvial systems have particular preservation (or taphonomic) characteristics. Indeed, floodplains, along with some other environments such as salt marshes and sand dunes, can have an excellent archaeological 'memory'. The information preserved, while providing exciting possibilities for the reconstruction of past living conditions, is controlled by a variety of physical factors; some are independent of human action, e.g. astronomically forced climatic change, and some are caused by human impact. If archaeologists are to use absence of evidence as data (which it is difficult to avoid), then all possible causes of a lack of preservation must be understood. This realisation was one factor in the rise of the research area of geoarchaeology (use of geological methods in archaeology) from the late 1970s onwards. Originally the result of collaboration and contact between archaeologists and geologists or geomorphologists (Davidson and Shackley, 1976; Rapp and Gifford, 1985), it has now become a recognised sub-discipline within scientific archaeology. This has occurred side by side with increased interest in other aspects of environmental and landscape archaeology and there is a large overlap between geoarchaeology and both environmental and landscape archaeology as illustrated by Hassan's (1979) 'components of geoarchaeology' (Table Intro. 1).
Introduction and the example of the Nile
3
A major objective of both geoarchaeology and broader environmental archaeology is palaeoenvironmental reconstruction, because data on changes in natural environments are of archaeological significance. It is for this reason that the first half of the book concentrates on physical and biological processes of alluvial environments which affect the formation, survival and bias of archaeological evidence from alluvial sites. While based on scientific methods, both the questions asked and the interpretation of data are as open to alternative paradigms as is the rest of archaeology (Clarke, 1972). Alluvial environments can also be said to have their own archaeology because throughout prehistory they have, despite great variability, been distinguishable from other environments in their resources and hazards. This has led in the past to somewhat grandiose claims being made for the sociocultural effects of alluvial settlement, including the development of hierarchical social systems (Wittfogel, 1957) and the rise, and in some cases fall, of urban societies such as Mohenjo-Daro (Jacobsen and Adams, 1958; Lambrick, 1967). Later work on the settlement patterns of the Tigris-Diyala floodplain based upon field survey (Adams, 1965; 1981) has enabled a more sophisticated level of analysis balancing environmental factors. Johnson's (1972) use of central place theory showed how important water transport networks were, relative to the optimum utilisation of usable land, for the location and pattern of settlements. However, the earlier over-simplistic and deterministic ideas illustrate a potential problem faced by most, if not all, geoarchaeological or environmental work. Namely, a neo-environmental determinism derived from correlations between environmental and artifactual evidence, as is caricatured by Burgess's suggestion that the cause of the apparent increase in the deposition of bronze artifacts in rivers, lakes and springs after 1500 BC was the development of a 'water-cult' related to increased precipitation and waterlogging (Burgess, 1980 cited in Shanks and Tilley, 1987). At the other extreme, some post-processual archaeologists relegate environmental evidence to an unimportant backdrop to the myths, traditions and experiences that condition human existence. This is environmental relativism as it implies that the particular nature of the environment will have had no significant impacts on that society or culture. This book lies, along with middle range theory, somewhere in between, taking as its starting point the specification of relations between environment, subsistence, technology and social ranking. The physical and biological resources of alluvial environments are seen as offering changing, complex and integrated sets of possibilities to human groups at costs which may or may not be acceptable when judged in their own terms. In this sense, the book follows a possibilistic view of the relationships between human behaviour and the environment. The environment not only filters and biases our initial data but must have affected the behaviour of our subjects in varied and subtle ways, including their perception of that environment. Alluvial environments are sensitive to changing climatic conditions and the Holocene has not been climatically constant
4
A lluvial geoarchaeology
(Kutzbach and Street-Perrott, 1985; Goudie, 1992). So the interaction between the climatic signal in alluvial stratigraphies and the cultural signal forms an underlying theme of this book. Since both the reaction of fluvial systems to external forces and human environmental modifications are spatially variable, this theme must be considered at the appropriate scales and we should be wary of making long-distance correlations or generalised statements. Part I of the book covers the principles underlying the physical and biological processes that operate in alluvial environments and the methods used in the study of these processes. Chapter 1 introduces floodplain evolution as this is essential for studies of environmental change and its relationship to human activity. Our knowledge of how floodplains have evolved is also dependent upon dating, which is covered in chapter 2. The interpretation of floodplain sediments and soils, covered in chapter 3, is a necessary part of all alluvial archaeology. This includes a basic understanding of the processes of river flow, sediment transport, erosion and deposition. An understanding of these fundamental factors is needed for the accurate interpretation of the fluvial 'jumbling' of artifacts. Most early prehistoric artifacts, and indeed sites, are reworked and disturbed. Hand-axes are little different from the normal pebbles carried by rivers, and sites may be little more than natural zones of accumulation of these 'pebbles'. However, because archaeologists attempt to interpret both axe form and artifact or site arrangement, the role of natural processes must be understood. In describing and interpreting an early prehistoric terrace site it is impossible for the archaeologist to ignore the palaeoenvironment and this normally means integrating the archaeology with some geomorphological interpretation of the site stratigraphy. In later prehistoric contexts, the floodplain may itself be seen, in part, as an artifact of human settlement within the catchment. In order to determine the real impact of cultures on fluvial systems the natural controls on those systems must be understood, otherwise there is a very real danger that the relationships between culture and environment are obscured. Floodplain excavations can provide a wealth of environmental data and chapter 4 deals with the methods of obtaining, analysing and interpreting archaeobotanical and archaeozoological data from floodplain sites. Part II focuses on applications of these principles and techniques to the interpretation of artifacts, sites and the relationships between human activity and the alluvial environment. Chapter 5 concentrates on those studies where the archaeology is confined to artifacts within the floodplain stratigraphy. This is most typical of Old World Palaeolithic archaeology and much of North American archaeology. One of the geoarchaeological advantages of the alluvial environment is the preservation of many sites through burial, and the study of buried sites and the history of alluviation is discussed in chapter 6. This history is related, in part, to land use, and floodplains can provide excellent information on land use history, as described in chapter 7. Indeed the floodplain is an important component of nearly all archaeological landscape
Introduction and the example of the Nile
5
studies. The last chapter explores the peculiarities of floodplains and low terraces for human occupation, and relationships between occupation and environmental change. This includes what T. Evans (1990) has called the asymmetry of human activities in the valley as compared with the valley sides. Can we assume that floodplain sites are rarely, if ever, simply sites that happen to be near rivers, i.e. how are function and environment related? Both positive and negative factors should be evaluated in any interpretation of floodplain settlement history, even if some of these factors are archaeologically invisible, such as perceived risk. While this book focuses on valley bottoms as blocks of landscape, they must be related to the record from the rest of the landscape. As Evans (1990) has stated: 'What wet sites share are preservation factors and similar environments. To divorce them from their [dry] regional cultural/ chronological context necessarily pushes their interpretation towards functional universals and environmental determinism.' This book is segmented so that those readers requiring the scientific/environmental background to alluvial environments can commence with Part I but those only requiring floodplain archaeology/?^ se can commence with Part II. The appendices are provided as elaboration of some of the more technical material covered in Part I, but material which is essential to practical or experimental work. An exhaustive global coverage is clearly impossible, so the aim throughout Part II is to take some of the more important or better-documented sites, largely from North America and Europe, and present them in the light of the theoretical discussion in Part I. The sites are united by the fact that, in every case, site interpretation has involved an element, sometimes large, sometimes small, of the reconstruction of a fluvial palaeoenvironment from stratigraphic evidence. In order to illustrate the overall theme of the book, i.e. how geological, geomorphological and hydrological methods can be used in alluvial archaeology, I propose to start with a classic example: the fluvial history of the Nile. The history of the Nile in recent times as well as in the more distant past is a complex mix of environmental, economic and legal issues (Howell and Allen, 1994), and because of its supreme importance to the Pharaonic civilisations it has been, and continues to be, the focus of much geoarchaeological study. This work has included a wide variety of methodologies and data sources drawn from subjects as diverse as Egyptology, geomorphology and climatology. An introductory example: the Nile The high Egyptian civilisation that emerged in the late fourth millennium BC was not only located on the Nile floodplain but was entirely dependent upon the Nile for its survival in an environment with virtually no rain and no fresh groundwater supplies. The classic work of Wallis Budge (1926) gives us some insight into the importance of the Nile in Egyptian life through its mythology and culture. Egyptians called the Nile Hapi which was also the name of the
Alluvial geoarchaeology
Fig. Intro. 1 Hapi the Nile God. The frog (a symbol of new birth) and the lotus plants on his head reflect the crucial role of the Nile in Egyptian agriculture as both appeared as a result of the flood. Adapted from Wallis Budge (1926).
God of the Nile. The God Hapi was depicted as a man with woman's breasts (Figure Intro. 1), indicating strength, and powers of fertility and nourishment. They celebrated the God at the festivals of inundation, and because of the central role of the river in the Egyptian economy the collection of accurate hydrological records began in the First Dynasty (c. 3000 BC) with the engraving of maximum annual flood height on a large stone stele. The first Nilometer (fixed recording device or structure) was cut into the rocks at Samnah in the Twelfth Dynasty (c. 2200-2000 BC). The Nile's behaviour formed the basis of the oldest Egyptian calendar, which divided the year up into three parts, one of which (Akhet) was the main flood season. The height of the flood controlled the amount of land watered and therefore the supply and cost of bread. If the Nile reached a height of 16 cubits food supplies would normally be sufficient to supply the population for a year and this was celebrated in the 'Wafa' festival. If the flood reached 18 cubits and the middle lands were flooded the 'Neirouz' festival inaugurated the new year's day of the agricultural year, and if 20 cubits was reached the 'Saleeb' festival was held, celebrating the flooding of the high lands (Evans, 1990). Successions of 'low Niles' always produced famines such as the seven-years famine which occurred in the Third Dynasty
Introduction and the example of the Nile
1
(c. 3190-3100 BC). The Nile was probably just as important in pre-Dynastic and post-Dynastic times, and so its Holocene history is an essential part of the archaeology of Egypt. The geomorphological history of the Nile has been reconstructed from studies of repeating sequences (cyclothems) of deposits in its delta (Stanley and Maldonaldo, 1979), sediments in the lower and middle valley (Butzer and Hanson, 1968; Adamson et al9 1980) and deposits in the swamps of Sudan (Williams, 1966). The archaeological importance of the alluvial environment has favoured interdisciplinary approaches as typified by the surveys of CatonThompson and Gardner (1932) and Sandford and Arkell (1939) and later by the work of Butzer. Butzer's (1976) 'cultural ecology' approach is quintessentially geoarchaeological, including studies of valley sediments, geomorphology and ecology. The valley which Butzer describes is bounded by limestone hills to the west and cliffs to the east (Figure Intro. 2). It is floored by Pleistocene sand and gravel and there has been a long-term tendency for the river to migrate towards the eastern side of the valley. During the glacial maximum (a 20,000-12,500 bp) the Nile was a highly seasonal braided river system with massive floods over four times larger than historical floods and periods of the year with no flow at all. This regime was caused by cold dry conditions in the East African Highlands and it produced a wide floodplain some 40 m above the present level between the first and second cataracts, where it corresponds with the area of the Sebilian culture. The Pleistocene period, which ended with overflow from Lake Victoria and increased rainfall in Ethiopia, was followed in the Holocene by permanent river flow and a period of extreme floods which caused incision of the river into its floodplain (Adamson et al, 1980). Up until about 5000 years ago higher rainfall than present in Ethiopia and a lack of flood storage in the slowly developing Sudd swamp resulted in centuries of high floods. These floods caused cycles of deposition and incision until Egypt and Nubia became more arid and less extreme when floods began to deposit the Arminna/Kibdi silts annually on the narrow Egyptian floodplain. Silt and sand deposition has produced natural levees and a domed cross-profile (Figure Intro. 3). Both sediments and surface features also indicate the presence of more channels in the past, some of which, such as the Bahr Yusef, still exist today. The existence of these former channels helps to explain the location of some Ptolemaic and older settlements such as Aphroditopolis and Heracleopolis. There is also documentary evidence of their existence (Strabo 17:1.4). In between these channels were large floodbasins which were flooded to a depth of about 1.5 m in a normal flood (Butzer, 1976). In its natural state the floodplain vegetation would have been evergreen forest of fig and acacia fringing the channels, and grassland or acacia-scrub savannah on the seasonally inundated flats. Palaeolithic settlements, such as those mapped by Butzer and Hanson (1968) in the Kom Ombo Plain, cluster on levees and riverbanks. This allowed easy utilisation of a variety of resources, including wildfowling in
Alexandria
AsyutW; EGYPT
Luxor